kentonv changed the topic of #sandstorm to: Welcome to #sandstorm: home of all things sandstorm.io. Say hi! | Have a question but no one is here? Try asking in the discussion group: https://groups.google.com/group/sandstorm-dev
frigginglorious has quit [Read error: Connection reset by peer]
frigginglorious has joined #sandstorm
frigginglorious1 has joined #sandstorm
frigginglorious has quit [Ping timeout: 258 seconds]
frigginglorious1 is now known as frigginglorious
frigginglorious has quit [Quit: frigginglorious]
michaeln3 has joined #sandstorm
michaeln3 has quit [Ping timeout: 260 seconds]
<isd> sigh I hate it when dev mode behaves differently than a built spk
<isd> Hit two of those today. Figured out one of them.
XgF has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
michaeln3 has joined #sandstorm
michaeln3 has quit [Ping timeout: 255 seconds]
codecowboy has joined #sandstorm
codecowboy has quit [Ping timeout: 258 seconds]
goonieb has joined #sandstorm
goonieb has quit [Quit: Leaving]
<ocdtrekkie> Apparently "GitHub Discussions" is coming: https://github.com/zeit/next.js/discussions
<ocdtrekkie> Curious if it'll be robust enough to replace Google Groups, or if it's just going to end up being a separate tab we move questions that aren't bug reports to.
xet7 has quit [Remote host closed the connection]
xet7 has joined #sandstorm
codecowboy has joined #sandstorm
codecowboy has quit [Quit: Leaving]
<ocdtrekkie> Now every test failure that was discussed recently has a tracking issue. #3239 and #3240 opened
<ocdtrekkie> kentonv: https://github.com/sandstorm-io/sandstorm-website/pull/317 is dated for today, if it is deemed worthy.
<kentonv> Published.
<ocdtrekkie> \o/
<ocdtrekkie> Published, I see a flaw with it that you made in your last blog post too.
<kentonv> ?
<ocdtrekkie> Subheadings in this one and "Shutting Down Oasis" are actually bigger than the page titles...
<kentonv> ah yeah, I sort of noticed that too
<ocdtrekkie> I am not sure if the best solution there is "don't use that heading level in blog posts" or "figure out how to ensure the CSS knows headings within posts should use smaller text". I also don't know how much people care.
<kentonv> meh. I don't think it's a big deal but if you feel like you want to reduce the heading level I'll be happy to re-push.
<ocdtrekkie> I'm not worried about in the immediate moment. Just observing.
<ocdtrekkie> Statistically likely a lot of people monitoring the Sandstorm blog are going to read it where I would: In an RSS reader interface that doesn't care.
michaeln3 has joined #sandstorm
<kentonv> doh... I wonder if there was a missing file
<kentonv> isd, are there any errors on the JS console when you see that?
<kentonv> (currently waiting for my own build...)
<isd> ocdtrekkie: want to add a label specifically for stuff like the broken tests? I would find it useful to be able to organize "crap we broke this" kinds of things.
<isd> Let me have a look
<ocdtrekkie> isd: Right now I just have them under sandstorm-dev (which is where I've been including things about building Sandstorm and tests)
<isd> Nothing in the console. Some errors in sandstorm.log; posted them on the issue.
<isd> Ack, brainfart. Yes, there's this in the js console: Error: There are multiple templates named 'changelog'. Each template needs a unique name.
<isd> this might just be a dirty working tree on my end actually; there's an uncommitted changelog.html file (that I didn't put there, weird...)
<isd> let me see if deleting that fixes anything.
<isd> Yeah, it does. Sorry for the false alarm.
<kentonv> heh, same thing happened to me when my build finished
Mitar has quit [Ping timeout: 260 seconds]
michaeln3 has quit [Remote host closed the connection]
michaeln3 has joined #sandstorm
TC01 has quit [Ping timeout: 260 seconds]
<ocdtrekkie> Fun fact, isd: Every time you force-push that branch, I get an email notification thanking me for noticing the clerical errors.
TC01 has joined #sandstorm
michaeln3 has quit [Ping timeout: 258 seconds]
<isd> Hah.
<isd> Yeah, because git can't tell that it's the same commit.
<ocdtrekkie> kentonv: Can you check the Sandstorm org's Actions and Packages storage budget?
michaeln3 has joined #sandstorm
<ocdtrekkie> I think I just found a potential billing issue.
<kentonv> where is that?
<ocdtrekkie> In my personal Settings, Billing, it shows me storage usage.
<ocdtrekkie> And in my case, I have 0 out of 0.5 GB of included storage used.
<kentonv> it says 0 of 2GB
<ocdtrekkie> The issue I'm predicting, is that abliss has his Action upload the build tarball, which is 88 MB, on every push and PR.
<ocdtrekkie> Interesting, I wonder what the Actions artifacts are stored in.
<ocdtrekkie> Or if it updates during a billing cycle.
<ocdtrekkie> To be safe, you might want to set up a $0 spending limit on storage for Actions and Packages.
<kentonv> the limit was $0 by default, but I increased it to $5
TC01 has quit [Ping timeout: 268 seconds]
<ocdtrekkie> If it actually bills you for any storage from Actions, we may want to remove that auto-upload of the built Sandstorm from the Action.
<ocdtrekkie> I'm not sure we really need it for anything. Presumably we only care about the built tarball if we're using it to do releases. Which then should probably not be an action that runs on every push.
<ocdtrekkie> The language suggests that running Actions on public repos is free, but that storing stuff for Actions is not.
TC01 has joined #sandstorm
<ocdtrekkie> Eh, maybe I'm wrong. The billing page is less clear. The help page says it's free for public repos.
<ocdtrekkie> $0 spending limit should be the way to go either way, we shouldn't be actually ever billing storage, and if we ever were, I'd want to fix it so we weren't storing stuff.
<kentonv> FWIW artifacts are automatically deleted after 90 days. But yeah it looks like this action runs pretty frequently so we should probably make it not upload a tarball.
<kentonv> also it looks like the tarballs it generates have build number 9999 which actually makes them dangerous to install, because they'll confuse the updater logic
<kentonv> FWIW I'm already paying $25/mo for the Sandstorm org account hence adding $5 seemed like no big deal to me.
<ocdtrekkie> Should both be easy fixes. Do we want to wait for abliss or just fix them?
<kentonv> but yeah I'll set it back to $0 since it seems like if we actually use more than 2GB, we're probably doing something wrong
<kentonv> eh I'm not too worried about it but if you opened a PR I'd probably merge it. :)
<ocdtrekkie> If only because playing with it might be fun...
<ocdtrekkie> Seems a little aggressive that it runs both on my commit to my branch and on my PR for the same commit... I canceled one manually.
<ocdtrekkie> ...aaand I have no idea why the other one failed.
<isd> I'm mildy amused by ocdtrekkie's aggressive updating of the ready-for-review label.
<ocdtrekkie> I am baffled by the build failures on the Action.
<ocdtrekkie> They don't seem to have a lot of rhyme or reason.
<ocdtrekkie> isd: I basically update it when reading notifications for my own reference, lol.
<isd> I figured. It's just kinda funny.
<ocdtrekkie> isd: tests/appHooks definitely fails on the Action, by the way.
<isd> ack, let me have a look
<ocdtrekkie> meteor npm install babel runtime or something like that.
<isd> Oh, yeah, that's a known issue.
<isd> I believe kentonv opened a ticket for it.
<isd> But the change needs to happen in meteor-spk, so I haven't dealt with it yet because updating that is not a completely trivial task.
<ocdtrekkie> ah okay. I guess it's failure just looked different than previous times I'd looked at the log. maybe because of the other fix that was merged.